What Does IT Operations Involve and How to Automate It?

1. What Are the Tasks of IT Operations?

Infrastructure: network, servers, operating systems, etc.

Environment management: development, testing, production environments.

Deployment: delivering applications or systems to various environments.

Monitoring: overseeing infrastructure, applications, or systems.

Alert response: handling and processing alert notifications.

Performance optimization: improving system and component performance.

High availability: upgrading single points of failure to high‑availability solutions.

SLA assurance: ensuring service availability and enabling auto‑scaling based on SLA.

The above tasks are derived from an operations management framework and are not exhaustive.

2. Current State of Operations

According to the 80/20 rule, about 80% of operations work can be handled through repetitive manual processes, while the remaining 20% requires case‑by‑case handling.

Automation can address the 80% portion, and the remaining 20% can be managed through multidimensional monitoring that collects and analyzes issues for further action.

3. Operations Management Goals

Given the current state, the primary focus should be on automation, which presupposes standardization and normalization. Effective automation is supported by visualization or web‑based tools, enabling optimization of a large portion of work.

Thus, the main objectives of operations management are standardization/normalization, automation, and visualization/web‑enablement.

Standardization can be tailored to actual operational conditions, while visualization/web‑enablement can be achieved with open‑source tools or custom web development.

4. Operations Automation

Key areas of operations automation include:

1. Server provisioning automation

Automating the entire lifecycle of a new server or virtual machine—from creation to delivery across environments—covers CPU, memory, disk, IP, kernel tuning, time sync, SSH hardening, firewall, client installations, and integration with CMDB, jump hosts, Zabbix, etc.

The ultimate goal is optimized, secure, and registered environments.

2. Environment definition automation

Two scenarios:

Small‑to‑medium companies where test environments share all systems and databases.

Large enterprises where each system requires an isolated test environment.

Automation is especially valuable for the second scenario, enabling rapid resource creation for demand‑side departments while maintaining isolation.

Regardless of scenario, the principle is to enforce isolation and reduce environment‑related errors.

3. Deployment automation

Deployment automation evolves through stages: scripts → batch SSH → automation tools → containers. Each stage improves scalability, usability, and efficiency, eventually addressing not only deployment but also speed and abstraction of underlying differences.

Note: This aligns with the DevOps mindset of improving speed after initial automation.

Post‑deployment, integration with monitoring (availability, performance) is essential.

4. Monitoring automation

Automation can add various monitoring dimensions, such as:

Automatic addition of availability checks (ports, URLs, etc.).

Automatic addition of log status checks (status codes, errors, etc.).

Monitoring automation also encompasses automated fault recovery (self‑healing).

5. Release automation

For smaller server farms, release automation must consider node removal and alert suppression, integrating with Nginx and monitoring APIs, e.g., smooth node removal in Nginx and API‑driven enable/disable of monitoring items.

5. Stages of Operations Automation

Understanding the maturity stages helps guide automation efforts:

1. Operational automation

Scripts or tools chain together manual steps, reducing manual effort but requiring constant adjustments for different scenarios.

2. Scenario automation

Tools make decisions based on external environment data defined beforehand, requiring extensive data collection and integration with third‑party systems (CMDB, network management, etc.).

3. Intelligent automation

Systems store operational data centrally (big‑data storage) and can analyze, decide, and act autonomously. The focus shifts to defining analysis strategies, maintaining the intelligent system, and intervening at critical decision points.

6. How to Implement Operations Automation

Before automating, recognize that enterprise architecture evolves rather than being designed from scratch.

1. Identify pain points

Classify recurring issues; automate what can be tool‑based and programmatic, reducing human intervention.

Whether or not a CMDB is used depends on the scale and change frequency of the environment.

2. Choose the appropriate maturity stage

Typical progression: manual support → online standardization → tooling → self‑service/platform automation. Select the stage that matches current business maturity; avoid over‑reaching.

For medium‑to‑large platforms, a CMDB and configuration system remain essential for accurate asset management.

7. Summary

Operations automation requires integrating various third‑party systems at each stage; a unified ESB (Enterprise Service Bus) can simplify these integrations, though it is not the only solution. The right automation approach aligns with business development stages.

Operations management aims for standardization, automation, and visualization/web‑enablement, evolving as automation maturity progresses.

In practice, many organizations have achieved scenario automation and are moving toward intelligent automation, freeing up roughly 80% of manual effort and prompting the next question: should operations evolve into an operations‑as‑a‑service model?